Capital Inflows and the Increase of FX Reserves

There is little doubt that massive capital inflows have caused the surge in FX reserves. Central bank statistics of Japan, China, Taiwan, Korea and Hong Kong, which hold the top five spots of FX reserves of the world, show that after the recovery from the Asian Financial Crisis, Asian economies have experienced rapid FX accumulation starting from early 2002 when the US dollar embarked on a two-year slump. Money continued to flow out of the US into Asia in the forms of huge trade and service surplus, FDI, portfolio investments and grey channels, etc.

A weakening dollar was the catalyst for such a development. Yet it should be recognized as a transitional rather than permanent feature. But the huge stockpiles of FX reserves by Asian economies should help them defend against future crisis or speculative attacks. Take Korea for example, it was one of the countries hardest hit in the Asian Financial Crisis. By the end of 1997, its FX reserves had dwindled dangerously to USD20.4bn. But six years later, they have soared to USD163.6bn or eight times the level at the darkest hour of the crisis. With a floating Won and refined debt management, Korea should be able to withstand speculative attacks several times of the scale back then. In 1998, Hong Kong spent USD9.5bn or HKD74.1bn of its USD98.1bn reserves to defend the Hong Kong dollar's peg. Nowadays, its reserves have grown up to USD123.8bn, protecting it from a much better position against a relapse of speculative attacks. China now commands the second largest war chest of FX reserves in the world at well over USD400bn, enabling it to deal effectively with any unforeseeable crisis in the future when it gradually opens up its economy. Therefore, there should be little doubt about the heightened financial security brought by the hefty accumulation of FX reserves.

The Differences in Impacts

Mr. Greenspan's recent warning against surging FX reserves originated from his belief that monetary policy goals could be compromised in the process of such accumulation. In China's case where the PBOC has decided to hike bank reserve requirements twice within one month seems to support such an argument. But after comparing the impacts from capital inflows and FX accumulations by Japan, China and Hong Kong, it can be shown that the differences could be significant because of the underlying financial systems and economic cycles. Orthodox economic theory may not be sufficient to explain some of the differences. Thus no general rules should be applied to all.

Japan has boasted the largest increase in FX reserves in the world on both the absolute and relative scales. Because the Yen is a floating currency, capital inflows should result in substantial appreciation of the Yen. However, the Japanese government has taken the stance of FX intervention to protect its export sector against a rising Yen. Statistics show that in 2003, Japan's FX reserves grew by USD230.8bn. And based on MOF's accounts, the FX intervention amounted to 20 trillion Yen for the same year. Using a medium exchange rate of 110 Yen per dollar, this mean that USD181.8bn or 90% of the increase in FX reserves came from interventions. Nevertheless, such large inflows could only produce 2.31% increase in its broad money supply M2+CD. Bank loans continue to decline by 3.52%. And real GDP growth for 2003 was 2.7%. One explanation is that Japan's fragile banking system has created a liquidity trap, rendering the inflows useless in stimulating the economy even under the zero-interest-rate environment.

But China's case begs to differ from such a conventional explanation. Based on international banking supervision criteria such as NPL, CAR and Liquidity Ratio, China's banking system is still weak by all accounts. Yet capital inflows have resulted in surging money supply, bank loans and economic growth. Such a case of a weak banking system performing admirably its financial intermediation function contradicts conventional wisdom and begs for alternative explanations. China's managed float foreign exchange system contains part of the answer. The RMB is allowed to float within a tight range only. The PBOC has to sell RMB actively to stabilize the exchange rate under constant inflows, during which money is created and money supply accelerates. Meanwhile, China's financial system is still relatively closed, with most of the capital accounts inaccessible. Under the WTO's opening timetable, China's state owned commercial banks are still under protection in order to win them time for reforms. Because the fund raising role of China's stock and bond markets has much to be desired, banks take up most of the financing slack as a result. Combined with a favorable rapid expansion cycle that fosters tremendous financing demands, beneficial interaction with inflows has occurred, further stimulating financing demands.

In Hong Kong's case, its banking system is one of the healthiest and most open not only in Asia but also in the world. Yet its exchange rate mechanism is the most stringent Linked Exchange Rate. Although capital inflows are of a less magnitude when compared to those of Japan and China, surge in monetary base has filtered through to broad money supply. Bank loans have also begun to recover as a result of economic revival. Due to the lack of sterilization by the HKMA, growth in Hong Kong's money supply exceeds that of FX reserves by a wide margin.

Through such simple comparison, it becomes obvious that due to structural differences in financial systems and economic development phases, capital inflows and the subsequent FX reserves accumulation have far different impacts, some of which prove to be difficult to explain by conventional theories. Therefore, analysis should be carried out and conclusion be reached on a case-by-case basis to avoid oversimplification. From such a point of view, China has its own rationale to resist the RMB revaluation pressure and conducts financial and banking reforms according to its own mandate.

The Costs of FX Reserves Accumulation

Asia's rapid accumulation of FX reserves does not come free even if it strengthens an economy's the financial position. The costs are related to what level of reserves is deemed appropriate, how to measure the management efficiency and what the opportunity costs are. As mentioned above, the differences could be significant.

By the end of 2003, based on central bank data and the US Treasury's data, it can be shown that 81% of Japan's FX reserves were invested in US treasury bonds, while Hong Kong's investment ratio was 49%, and China, with the second largest FX reserves in the world, had only 37% of its money in US treasuries. If taking into account of the USD45bn used for capital injections into two state-owned banks, China's ratio could be even lower.

Opportunity costs may explain such vast divergence in the holdings of US treasuries. In order to keep local currency from appreciating under constant inflows, a central bank has to keep selling local currency and buying the dollar. An interest rate arbitrage is performed during the process if local currency interest rate is lower than the dollar's interest rate. In Japan, the zero interest rate policy begs for such an arbitrage. And in Hong Kong, ample interbank liquidity as a result of large inflows drives down market rates to levels well below the US counterparts. As a result, Japan and Hong Kong incur basically little such opportunity costs in their FX reserves accumulation. On the other hand, China has to weigh in such costs, as the RMB interbank rates are higher than the dollar rates. Therefore, China has less appetite for US treasuries as they offer higher safety but lower yields. Instead, US agency papers or investment grade corporate papers may be what China prefers to make up for the interest rate differentials. The cost of Japan selling the Yen is practically zero, therefore it does not mind buying the lower yielding US treasuries. And Hong Kong strikes in the middle ground dictated by its rate consideration.

Unlike the HKMA, Bank of Japan and the PBOC have practiced issuing short-term papers to sterilize the FX intervention and rein in the growth of domestic money supply. According to the yield curves of the four economies involved on April 14th, again Japan and Hong Kong are in a position to benefit the most from such sterilization, as their yield curves are lower than the US. China's higher yield curve boosts the cost of sterilization, making it less inclined to invest heavily on US treasuries. It can be safely assumed that other assets carrying higher rates or other higher yielding USD bonds are more suitable choices for China.

The Risks of FX Reserves Accumulation

Central banks are facing the foreign exchange and interest rate risks after building large FX reserves. As the dollar remains the reserve currency of the world today, and dollar assets still account for the largest proportion of any official FX reserves, dollar depreciation should not affect the nominal value of the reserve assets. Instead, if no hedging is performed on assets denominated in other currencies, dollar depreciation will result in exchange profits from these assets. Yet during the past two years, central banks have been getting more wary of a weakening dollar and diversifying more into other assets. The motives behind include the exchange profits in case the dollar depreciates more and the intent to limit the deterioration in the purchasing power of dollar assets. If the US dollar is deemed to have bottomed and will rebound in the future, we are likely to see central banks renew their preference for dollar assets.

Interest rate risk is the single most serious risk facing FX reserves, as dollar assets account for the majority of any FX reserves and dollar bonds account for the majority of dollar assets. This is consistent with rules governing the risk management of FX reserves and thus there is practically no exception. But recently, dollar bonds have been exposed to tremendous interest rate risks. In the last month or so, a series of strong economic data have completely reversed the market's view on US interest rates. Rate hike expectations gather steam and the US market rates soared. By April 15th, the Total Return Indices of US Domestic Bonds complied by Merrill Lynch show that total return of holding US treasury bonds YTD is 0.233%, and that of Agency bonds 0.425%, and investment grade corporate bonds 0.533%. In other words, if market rates increase further, investments in these bonds will turn from being profitable to losing money on the year. This uncertainty will no doubt continue to haunt FX reserves investments.

The US Federal Reserve has held rates at historical low levels for a prolonged period in order to stimulate the economy, fend off various crisis and combat deflation. Such a stance makes carry trade the fashion of the bond market. Many commercial banks, investment banks, brokerage houses and hedge funds engage in these trades by borrowing funds at low short-term rates and investing them in longer term US treasuries, easily pocketing the inertest rate differentials as profits. By the same token, the practice of Asian central banks intervening in the FX market and then reinvesting the proceeds into the US treasuries is just another form of carry trade. This is because either Asian currencies carry rates lower than the US or their short-term papers used in FX sterilization carry lower rates as well. However, if the US short-term rates are raised rapidly, carry trade in the US will see a relapse of the 1994 debacle. And if Asia's interest rates rise quickly for some reason, the carry trade embedded in the FX reserves accumulation will also be severely tested. Take Japan for example, as long as it can maintain a zero interest rate environment, it will continue to benefit from inflows and rising FX reserves because the carry trade is profitable. But this phenomenon cannot last forever. Once the old balance is destroyed, the impacts could be broad based and severe, which happens to be the biggest risk as of now.